The Heuristic Approach to the Selection of Experimental Design, Model and Valid Pre-Processing Transformation of DoE Outcome
Вантажиться...
Дата
2014
Автори
Pietraszek, Jacek
Goroshko, Andrii
Назва журналу
Номер ISSN
Назва тому
Видавець
Trans Tech Publications
Анотація
The typical problem in the design of experiments methodology is to avoid the risk of obtaining predictions outside of ranges with a technological or physical sense. Such a situation occurs very often if a simple linear regression or another unbounded one is involved as a predictive model. The possible solution is to provide mappings of the outcome from the range of interest into the full range from negative to plus infinity before the linear regression is applied. The intermediate values obtained from the unbounded predictive model are re-transformed into the physical domain by the inverse transformation. The key issue is to decide if the mapping is required and –subsequently – what mapping is necessary. The author proposed a simple heuristic solution supporting this decision and tested such a solution in some examples described in this paper.
Опис
Ключові слова
heuristic decision, design of experiments, experimental design, DoE, fractional factorials, response surface, logistics regression
Бібліографічний опис
Pietraszek, J. The heuristic approach to the selection of experimental design, model and valid pre-processing transformation of DoE outcome / J. Pietraszek, A. Goroshko // Advanced Materials Research. – 2014. - Vol. 874. – P. 145-149.